Salary After Cost of Living: 48 US Software Markets, Ranked

San Jose has the most expensive housing in the United States. It runs at 211.9 on an index where the national average is 100, so more than double.

The same federal index says San Jose overall is 10.4% expensive.

Both numbers are published by the same agency for the same year, and the distance between them is why every cost-of-living calculator feels wrong to anyone actually paying Bay Area rent. Housing is one component of the overall index, averaged together with groceries, utilities, healthcare and services, and those barely move by geography. Average them all and the component that differs gets diluted until it nearly disappears.

That gap is not a San Jose curiosity. Across 387 US metro areas the overall price level runs from 83.6 to 115.6, a spread of about 1.4 to 1. Housing runs from 42.3 to 211.9. Five to one.

So I converted every salary in the federal wage survey into national-average dollars to see what a paycheck is actually worth once local prices come out of it, and the results cut in directions that surprised me.

What was measured

Two federal datasets, joined on the metro area.

The first is the Bureau of Labor Statistics wage survey, which publishes the 25th, 50th and 75th percentile of pay for each occupation in each metro area. The second is the Bureau of Economic Analysis Regional Price Parities, which measure, in the agency’s words, “the differences in price levels across states and metropolitan areas for a given year,” expressed as a percentage of the national level. San Jose sits at 110.4, meaning prices there run 10.4% above the national average. Austin sits at 98.1, 1.9% below it.

Dividing the first by the second converts every salary into national-average dollars — what the number would buy if you spent it at average US prices. That is the column this article ranks on. The full table covers 47,380 occupation-by-metro cells across 387 metro areas and 368 occupations, and it is free to download with the column reference under CC BY 4.0.

For the ranking below I used software developers, the largest well-paid occupation with markets in every region, and kept only the 48 metro areas employing at least 5,000 of them. Below that threshold the sample gets thin enough that a single large employer moves the median.

The expensive cities mostly stay ahead

Given all that, the first finding argues with the thing everyone says. The standard line is that a coastal salary is an illusion, that it evaporates the moment rent is due. Mostly, it does not.

San Jose pays a median of $213,110 against Austin’s $134,120, a gap of $78,990. In national-average dollars those become $192,994 and $136,765, and the gap becomes $56,229. Seventy-one percent of it survives. San Jose remains the highest-paying market in the country for this job by a distance, and no amount of price adjustment brings it back to the pack.

Austin does gain: its prices are below the national average, so the same paycheck stretches further and it climbs from ninth to eighth among the twelve largest markets. But the reordering is modest, not a reversal. Adjusting for prices dents the coastal premium. It does not delete it.

The full ranking

All 48 metro areas that employ at least 5,000 software developers, ordered by what the median buys at national-average prices. The housing column is the same index restricted to housing, and it is the one worth reading twice.

The sharpest reversal in the table is Miami against Huntsville, Alabama. Miami pays software developers $4,570 more; in national-average dollars the Huntsville developer ends up $21,412 ahead. Denver pays $12,620 more than Kansas City and ends up $4,974 behind it. Miami pays more than Pittsburgh, Kansas City, St. Louis and Cleveland, and finishes 47th of 48. Huntsville, which pays less than every market in the top twenty by list price, finishes 13th. Seattle slips from third by list price to fifth, and Los Angeles from eighth to tenth.

The pattern is not “cheap places win.” Charlotte, Durham, Huntsville and Kansas City rise because their price levels are genuinely low while their software salaries are not far off the national middle. Miami falls because it charges coastal prices on a salary that is not coastal.

Smaller markets sit outside this table because the wage sample gets thin below 5,000 developers, but they are in the dataset. Omaha would land around $136,240 in national-average dollars, ahead of Denver, Chicago, Dallas and Miami. Las Vegas would land near $129,661.

Which of the two numbers applies to you

The index that says “San Jose is 10% expensive” and the lived experience that says “San Jose housing is double” are both correct. They are describing different things, and the housing column in the table above is there so you can read them separately.

Which one applies to you depends on a fact the data cannot see: whether you are about to start paying local market rent, or already own, or are moving in with family. For someone facing a new lease in a hot market, the all-items adjustment understates the hit. For someone whose housing costs are fixed, it overstates it. That is why the dataset ships housing as its own column instead of burying it inside a single cost-of-living score.

What this does not change

One thing worth stating precisely, because it gets argued about a lot.

Adjusting for local prices does not change how much pay varies inside a job. The gap between the 25th and 75th percentile in one occupation in one metro divides by the same price index at both ends, so the ratio comes out identical to the last decimal. The spread of an occupation is what tells you whether there is anything to negotiate at all — which jobs have room to negotiate is a question about the shape of the distribution, and cost of living leaves that shape untouched.

Local prices change what a number is worth to you. They do not change how much the market around you varies.

Four limits, stated plainly

The two sources are a year apart. The wage data is May 2025; the newest price parities are 2024. Price levels move slowly relative to the gaps measured here, so the direction of every finding holds, but the exact dollar figures carry that seam.

Price parities are spatial, not temporal. They compare places within one year. They say nothing about inflation, and a metro at 110 is not “10% more expensive than last year.”

The index is an average basket, not your budget. It assumes roughly average consumption. If your spending is unusually housing-heavy or unusually housing-light, your personal adjustment differs from the published one, sometimes by a lot.

Base pay only, and no taxes. The wage survey excludes equity, signing bonuses and annual bonuses, which are largest in exactly the markets that top this table — so the top of the ranking is understated relative to total compensation. State and local taxes are a separate layer the price index does not touch, and they run in the same direction as the ranking in Texas and Florida and against it in California and New York.

Using this in an actual conversation

The practical use is narrower than “move to Huntsville.”

If you are comparing two offers in two cities, convert both to the same dollars before you decide, and then look separately at housing, because that is the component that will differ most and the one an averaged index smooths over.

If you are negotiating a single offer, the price level of your city is not leverage. Nobody funds a raise because their employee’s metro is expensive; the range was set by the market for the role. What the ranking is good for is knowing whether the number in front of you is genuinely strong or only nominally strong, which changes how hard you are willing to push and what you are willing to trade for. When you get to the actual ask, how much to counter is a question about the role’s range, not the city’s prices.

And if you are considering a move for money, the gap you care about is the price-adjusted one, minus the housing difference you will personally face, minus the tax difference. This article closes the first of those three.

The full dataset, all 368 occupations across 387 metro areas with nominal pay, price parities, housing parities and price-adjusted figures, is free to download and reuse, alongside the pay-band table it was built next to.